A lot has happened since I last wrote about the OpenDocument vs. Microsoft file formats drama that is clearly turning out to be one of the most important beachheads in the computer industry. Shortly after Microsoft announced plans to submit its formats to Ecma International for ratification, and then subsequently to the International Organization of Standardization (ISO) for its imprimatur as an international standard, the Redmond, WA company released a special covenant not to sue that was met with the satisfaction of Larry Rosen, one of the open source movement's leading attorneys.
But in the days the followed, the series of announcements were picked a part by critics who took issue with several possible gotchas. Among them, the assignment of the new legal language to Microsoft's current formats instead of the ones that Microsoft is trying to set as the standard. Although the language raised concerns that developers might not be as free to deploy Microsoft's future XML schemas as it is the current ones, Microsoft quickly diffused the bomb saying that the language will apply to the more forward-looking formats. I believe them. If the company doesn't keep its word, it will be crucified.
Then, there was the conformance proviso that basically guarantees that Microsoft won't sue developers as long as they conform with 'the standard' in its entirety. Rosen flagged the proviso as well, but told me via e-mail that the covenant not to sue was such a significant olive branch for Microsoft to offer -- practically a bitter pill for the software giant to swallow -- that taking issue with the conformance proviso was the equivalent of sweating small stuff that developers shouldn't be sweating. Wrote Rosen, "And the scope of their patent covenant, even though it is limited to 'conforming' software products, is sufficient to allow open source implementations that can read and write Office 2003 documents." Although one interpretation of the proviso is that Microsoft will sue developers of non-compliant implementations, that's actually leaping to a conclusion that can't necessarily be drawn. No Microsoft legal document states that. So, it's not clear what exactly would trigger a Microsoft lawsuit.
Another question that surfaced in the blogosphere had to do with the fact that the OpenDocument Format is already before the ISO for ratification. Several bloggers, most notably from Microsoft's competition, cited a problem with this asking how the ISO could be expected to put its imprimatur on two standards that essentially address the same problem. The concerns led me to ask (in one of my prior blog entries) "Knowing that an alternative to ODF could land in its lap in the coming months, will the ISO put the brakes on its current deliberations over ODF until it can vet the two side by side? Or, supposing ODF does get the ISO's imprimatur; under what conditions might the organization also give its imprimatur to Microsoft's file formats?" If you read on, you find out why. I found the answer.
The final concerns had to do with Ecma International -- a 45 year old European consortium that doesn't require the specifications it puts its imprimatur on to be available on a royalty free or unencumbered basis. Instead, the patent holders whose intellectual property might be included in an Ecma ratified-specification are at the very least required to license the relevant patents on a reasonable and non-discriminatory basis (otherwise known as RAND licensing). Most people are comfortable with the non-discriminatory part. But as you'll see in a minute, even the Secretarie General of Ecma admits there's no official definition for "reasonable" which means that vendors whose patents are a part of an Ecma specification can charge anything (including nothing at all) as long as they charge the same price to everyone (the non-discriminatory part).
Taken in combination -- the coverage of current formats but not the new ones, the conformance proviso, the confusion over two standards at the ISO, and Microsoft's selection of Ecma (an organization with a RAND-is-as-tough-as-we-get policy) -- Microsoft's opponents saw enough of a confusing labyrinth of related events and announcements that they've grown suspicious that some nefarious plot is afoot. If you read any of the anti-Microsoft blogs on the issue, it's pretty clear that if the gotcha isn't found, that once the dust settles, the OpenDocument Format will have been filibustered into oblivion, Microsoft's formats will retain their position as the de facto standard (with the added dubious distinction of de jure standard as well), and somehow, for developers to develop office suites that are 100 percent file format compatible with Microsoft Office 12 and whatever comes after that (from Microsoft), they'll still have to go through and perhaps pay Microsoft.
[Update: Based on some comments to this blog, I want to make it clear that this is a summary of the concerns that have appeared in the blogosphere and not an opinion held by me. As you'll see from my questions, I'm less interested in finding the gotcha and more interested in understanding Ecma's process and how Microsoft can insist on conformance when the new standard is supposed to be multiparty-stewarded. Barring the conformance question, openness from an implementation point of view is not an issue. That's because Ecma already allows for royalties to be attached to its standards. So, we can't look to Ecma to enforce the openness. For that, we have to fall back to Microsoft's legal language which is actually more open than what Ecma requires. That said, I think it's worthwhile to summarize the concerns in language that most can understand.]
One simple possibility for example is that once Microsoft's specification or a derivative thereof gets the Ecma and ISO imprimaturs, that Microsoft will use its current market presence to turn an extension of that specification -- one that developers don't have unencumbered access to -- into the de facto Office format standard. Much the same way Microsoft's claims its Internet Explorer is "based" on open standards (it is, but it also includes support for Microsoft-specific proprietary extensions), it isn't difficult to envision a claim that Microsoft Office is based on international standards and for the de jure standards to become marginalized in the process.
[Update: Microsoft's .NET is also an example of this, and perhaps a better one. The Common Language Infrastructure (the CLI) is a core piece of .NET that Microsoft submitted to Ecma for standardization. For all intents and purposes, .NET which existed before the CLI was carved out and submitted to ECMA, now represents a proprietary extension to Ecma's CLI standard. Please note however that the existence of prior examples such as these is in no way a guarantee that Microsoft will follow the same pattern with the Office XML file formats. In fact, chances are it will not. Especially given how many people watching like hawks and waiting to say 'I told you so.']
Possibilities such as these are what concerns some that Ecma and the ISO are being unwillingly played as pawns in the company's larger struggle to somehow maintain the control it has over the desktops of so many (while keeping substitutes at bay). Microsoft's selection of Ecma comes as no surprise. On the surface, Ecma is not only a European consortium -- Europe being one of the regions in the world that has grown increasingly intolerant of Microsoft -- it's also a multiparty consortium whose structure is similar in nature to OASIS (the consortium that put its imprimatur on the OpenDocument Format). One criticism that Microsoft has faced regarding its formats -- especially from the Commonwealth of Massachusetts -- is that Microsoft is in control of them. The state has been clear that it prefers the standards it settles on to be stewarded by multiple parties and not just one company. By turning its formats over Ecma whose ratification process involves technical committees that involve multiple parties, Microsoft will be able say that its file formats -- at least the ones ratified by Ecma -- are multiparty stewarded (which in turn should satisfy organizations like the Commonwealth of Massachusetts).
Another reason Microsoft probably selected Ecma is that the sooner Microsoft's XML formats get the imprimatur of the ISO, the better chance they have against the OpenDocument Format. Ecma is one of the only two consortia in the world (that aren't officially affiliated with a government) that can put a specification on the ISO's fast track: a special track that not only speeds up the ratification process, but, as you'll find out from my interview with Ecma's Secretarie General Jan van den Beld, one that's almost always guaranteed to end in success.
The selection of Ecma along with Microsoft's statements do in fact raise some interesting questions -- some of which I asked van den Beld in my telephone interview of him. For example, Microsoft has said that it expects the formats in question to have the imprimaturs of both Ecma and the ISO within the year. One good question is how, if the process is truly stewarded by multiple parties, Microsoft can set such an expectation. Presumably, when a process is multiparty stewarded, none of the involved parties can really set that expectation. Technical committee deliberations at just about any multiparty consortium typically involve heated debate and competing interests that can take an unpredictable amount of time to resolve. Another question has to do with the events of this week where Microsoft not only formally requested that Ecma establish a technical committee to oversee the stewardship of the XML formats the company was submitting to the consortium, but included language that appeared to request that the TC approve the submission without any real deliberation. In suggesting a scope for the technical committee, the Microsoft request proposed:
The goal of the Technical Committee is to produce a formal standard for office productivity applications within the Ecma International standards process which is fully compatible with the Office Open XML Formats.
Almost as soon as the proposal was published, the "fully compatible" part set off alarms throughout the blogosphere. Instead of suggesting that the goal of the technical committee would be to produce an open XML-based standard for office productivity applications that can use Microsoft's submission as the starting point for its deliberations, the proposal at the very least appears to angle for a rubber stamp approval of Microsoft's submission. Now that Ecma has accepted the proposal and officially formed the TC (TC45, based on the age of Ecma International -- 45 years old), one question I have is whether or not that very narrow goal statement imposes any restrictions on the latitude that the TC has to change the specification. If it doesn't, it certainly draws the multiparty nature of the TC into question as far as organizations like the Commonwealth of Massachusetts are concerned. If the other parties can't change the specification, is it really a multiparty specification, or is it still pretty much controlled by Microsoft? I contacted Microsoft to get its take on the question on Wednesday (12/7) but have yet to hear back.
If Microsoft was looking to make sure that the conformance proviso stayed with the specification, it may have given itself no choice but to see rubber stamp approval of its specification without additional contributions from other members of the TC to the point that the end result could be considered a derivation. One question I had, prior to the text of Microsoft's proposal to Ecma being published, was how it could insist on conformance with a specification that would technically no longer be within its control once it was turned over to Ecma. In other words, who is Microsoft to insist on conformance when the final specification, if it's truly multiparty stewarded, turns out to be a derivative of Microsoft's original submission because of the contributions made by the other parties?
Unless all of those other parties (the ones making the contributions that resulted in the derivation) agree that complete conformance is a requirement, then I don't see how Microsoft's covenant not to sue can be predicated on complete compliance with the Ecma-ratified specification. Put another way, it's basically like saying you can use our intellectual property only if you use all of the other contributors' IP too. If you don't, then not only can't you use any our contribution, you can't use any of the other contributors' IP either. I'm not a lawyer, but what place is it of Microsoft's to set the terms of my usage of others' intellectual property? How about if I do a partial implementation of the Ecma-ratified specification that only includes the IP of the non-Microsoft contributors and that doesn't include any of Microsoft's IP? It would seem to me that only the owners of that intellectual property can set the terms of the usage of their IP. If Microsoft can impose such terms on its intellectual property as well as that of the other contributors, then it once again calls into question whether or not the specification is truly multiparty-stewarded. On the other hand, if Microsoft changed its terms to say that I only have a license to 'practice' its patent for the purpose of using it in my implementation (partial or complete) of the Ecma-ratified specification and that I cannot use it for anything else (eg: to develop some other unrelated product or technology), that's OK. This is not uncommon in the standards business.
This is one reason it seems to me that the conformance proviso and multiparty stewardship of an Ecma technical committee are incompatible with each other. Again, I sought comment from Microsoft but the company still has not gotten back to me yet. When it does I'll let you know. In the meantime, these questions turn the spotlight onto Ecma's processes. Is Ecma set up to guarantee multiparty stewardship or are there ways for a single contributor to sieze control of its process? To get to the bottom of these and other questions, I interviewed the consortiums Secretarie General Jan van den Beld. van den Beld has been either the #1 or #2 person at Ecma International since it was first founded in 1961. In our conversation, he was surprisingly candid with me, revealing that he's completely aware of how (1) Ecma's RAND licensing paves the way for patent holders to encumber the consortium's standards and that (2) given the leverage patent holders seek to gain through an Ecma imprimatur, its only natural for Ecma to actually ratify multiple standards for the same thing (as it has done in ratifying two separate standards for rewritable DVDs). van den Beld says the same philosophy goes for the ISO as well. Here's the link to the interview.
In addition, I asked IBM and HP, both of which are members of Ecma if they'll be participating in the new technical committee. I also asked Sun, which is not a member, if they would be joining in hopes of participating. Sun has no plans to join as of this time. Here are links to the responses from IBM and HP.